WMO Launches First AI Nowcasting Project in Asia with Shenzhen Kick-off Meeting
The World Meteorological Organization (WMO) has launched its first Artificial Intelligence for Nowcasting Pilot Project (AINPP) for Asia with a regional nowcasting intercomparison activity that began on May 9, 2025, in Shenzhen, China. The initiative, organized by the National Meteorological Center of China and the Meteorological Bureau of Shenzhen Municipality, brings together experts from WMO, the China Meteorological Administration, and meteorological departments from Japan, Republic of Korea, and other countries in both online and in-person formats.
Yuki Honda, Chief of Earth System Prediction at WMO, described the launch as "an important milestone in the ongoing efforts of countries to advance modern warning and forecasting technologies." The project aims to strengthen international cooperation in AI-powered meteorological services and directly supports the UN's "Early Warnings for All (EW4All)" initiative.
As WMO's first AI pilot project, AINPP promotes multi-disciplinary and multinational collaboration between public and private sectors. The project convenes international meteorological and AI experts, including advanced AI practitioners and representatives from developing countries, to develop, evaluate, and implement AI-based nowcasting models.
Participants at the launch of the WMO Asian Nowcasting Intercomparison project in Shenzhen
Project Objectives
The intercomparison activity, a key component of AINPP, aims to:
- Provide a collaborative framework for nowcasting activities in the Asian region
- Facilitate evaluation and comparison of nowcasting products
- Prepare for the implementation of AI nowcasting products in developing countries
Participants include meteorological experts from Thailand, Vietnam, India, UAE, Saudi Arabia, Hong Kong Observatory, and AI specialists from Guangdong, Guangxi, and private sector companies including Microsoft. The project has collected 13 AI-based forecast technologies related to lightning, precipitation, and radar echo.
Why Shenzhen?
Jun Yu, Regional Officer of the WMO Regional Office for Asia and the South-West Pacific, noted that Shenzhen was selected for this activity because of its frequent typhoons and rainstorms, which provide the necessary climate conditions for the project. Significantly, Shenzhen has previously hosted the WMO RA II Pilot Project on Public-Private Engagement for Smart Meteorological Services in Mega-Cities (PPE-SMSC) jointly proposed by the China Meteorological Administration (CMA), the Hong Kong Observatory (HKO), and the Macao Meteorological and Geophysical Bureau (SMG).
Shenzhen's AI Applications in Meteorology
The Shenzhen Meteorological Bureau has already made significant progress in applying AI to early warning services. Examples include:
- Development of the "Zhiji" regional forecast model with Huawei Cloud, achieving precise 3km resolution forecasts with improved stability and accuracy for typhoon path prediction
- Partnership with Harbin Institute of Technology (Shenzhen) to develop the AI-based short-term nowcasting model "Zhitong," which extends effective warning time from 1 hour to 3 hours
The Shenzhen Meteorological Bureau plans to implement relevant forecasting technologies during this year's flood season, continuously testing and validating them to enhance meteorological service support. Moving forward, the bureau will collaborate with local high-tech enterprises and universities to develop AI models adapted to nowcasting scenarios, focusing on solving prediction challenges for hazardous weather.
Future Plans
As the center for the Asian nowcasting intercomparison activity, Shenzhen will establish a data sharing platform and conduct cross-regional product verification and operational applications, actively participating in the "Developing Countries Assistance Program" to provide more Shenzhen solutions for regional disaster prevention and mitigation.
Source: Meteorological Bureau of Shenzhen Municipality, May 10, 2025